Howie Liu on Airtable's Early Days, Scaling, and AI

on year. And we had like, I think literally like 25 people and we didn't even—

25 employees.

25 employees total. And we could have just kept it close to there because again, like we didn't need to go and like hire large armies of salespeople. We could have automated a lot of support stuff. And, you know, obviously you want to continue maintaining and evolving the product, but like we could have done that on a skeleton crew. So there is an advantage of PLG for sure. It's like very efficient. I think the disadvantage is that, A, sometimes you lack clarity of who the buyer is because like you're just getting it, you know, they're coming to you. And, you know, PLG companies can be verticalized. So many of you may have verticalized PLG companies where it's a specific type of customer in a specific use case and industry that's coming to you organically. But I think more often than not, a lot of the PLG companies are also more diffuse in their top of funnel because they're not going and targeting the customers they want.

So on that point, let's take the mic to Julia because Julia, you asked about how to identify enterprise leads from the self-serve Did you want to elaborate on that and introduce yourself too for people?

Wow. Yeah, it's great to meet you. I'm Julia. I'm the co-founder and CEO of Kapwing, which is— we're trying to be a fully online video editor, sort of like a Figma but for video creation. And yeah, I'm so excited to hear the next thing you're about to say about what happened after the Series C because I feel like we're in the same place right now. But yeah, I'm interested in sort of hearing some about when you hired your first salesperson and how you think about whether you did that at the right time. And also some about how you nurture, or at least how at the early stages you identified your earliest enterprise-y kind of customers and developed them from a small team using Airtable, 5 to 10 people or whatever, to all of a sudden being like, oh, there's pockets of teams, and then selling to the enterprise.

For sure.

So this is a very complicated topic, so I'll try to keep it simple. I think one is There is an often fundamentally different value prop you have if you're PLG, because the end users or maybe a team manager is coming in and self-serving into your product. They're deciding this product is for me, I'm going to get value out of it, I'm going to try it, I'm going to use it. But often with enterprise sales, to make the economics work, depending on how much you're charging per team or per individual, you can't go and have salespeople efficiently go out and sell individual people. So you have to sell a bigger chunk of seats at a time. And often that means you're going to a more senior buyer. So instead of like, you know, maybe this video production person or, you know, team manager, you want to go to the head of production or post-production at a company. And that person may care about very different things than what the individual cares about, right? And this is the whole tension between like historically top-down sold enterprise software, which end users hate. I won't name any names since we're on camera, but you can imagine that, you know, software products that you don't enjoy using in enterprise, you know, but like the reason why they still succeed is because like they serve the needs of the buyer, right? So if you're selling to the CHRO or the CIO or the CRO, they care about different things like having really good visibility over your sales pipeline or like, you know, having like a master, you know, kind of record of the employee and everything happening to them more so than like the experience for the end user, right? And so I think like the first point is, you know, I think you'll often see with PLG companies this like difficult transition between, okay, we were really good at serving the individual or small team unit with our product, but as we try to go and staff up enterprise sales, they're trying to go and sell higher, but we don't have a value prop that directly serves the needs of that more senior buyer beyond your teams like us, right? And I think one thing that really changed, at least we saw change between like, call it 2021, '22 ZIRP era buying behavior to now, is that in that ZIRP era, especially like post-COVID or during COVID I think a lot of companies were just willing to like spend money on everything. Like budgets were really, you know, kind of lax in IT. And there was like, you know, kind of this catering to the employee demand of like, whatever tool you want, we'll, we'll basically let you foot the bill, right? And of course, like over the past couple years, there's been a demand reset. Part of that has been about like consolidating tool sprawl. Like, you know, every company's like, we have too many tools. They're performing duplicate purposes. Some of them are not actually adding business value. Let's cut them down. Some of it's about like, let's, you know, rationalize the spend. Like even with hyperscalers, let's make sure we're, we're like making every seat or every compute cycle count. But The dynamic has switched so that it's harder to make the case of, hey, like, you know, VP of production, buy our product because 50 different people in your group like us and the other 50 people would like us as well. It has to be like, and we give you this X value that you care about. Maybe it's like standardized production workflows or standardization of brand assets or something, right? That usually is more about global visibility or standardization or some benefit across the entire org.

So Julia, how has this landed, or do you have follow-ups on this? Because you also had a question, when did they hire their first sales reps, or do you want to elaborate?

Yeah, no, I mean, this is definitely helpful. Yeah, I'm really interested in sort of hearing about at that point when you were hiring your first salesperson and they were communicating this value proposition. What were some of the things you learned? What were some of your insights at the time about ways you did that well and mistakes that you made that we could avoid?

So what I'm trying to disentangle also is that when we hired our first salesperson, this was back in 2017, '18. And that was a very different time where we could get away with— in fact, the first era of sales for us was more about demand capture selling, which is basically how I describe this earlier version of selling where you just go into the company and say, like, you have a bunch of people already organically using Airtable. They're like, there's more of them piling on. And wouldn't you like to buy, you know, seats in bulk, both to get some volume discounting and also to administer them all? But like, the trend line is there and people, you know, there's momentum. So buy on momentum, right? Basically. And the answer to that was generally yes from the buyer at the time. So We were able to go and just hire more salespeople that did more or less some form of this where like the sales pitch was going in showing a graph of usage at the company organically growing and then saying like, you know, we, we give all these teams collaboration value and make them more productive. And here's quotes from the teams now buy. And they would.

Right.

And what I'm saying is I don't think that same playbook would work as well today for you because it's a different time. And we've also adjusted, um, you know, as well. And I think what we're now doing a lot more of is coming in and selling to a more strategic benefit. So as an example, like we'll sell the CPO of like a very large enterprise organization, like a Fortune 500. And for them, the benefit is not like, okay, Airtable is a very usable, friendly, easy to flex product, but rather you can standardize all product operations across every product group that you have and have a master view of all of the product initiatives and how they're tracking and, you know, how they roll up into like your company-level goals, right? And that actually required some product, uh, work as well, right? Like we had to build the ability to kind of roll up across multiple instances of Airtable. We built higher-scale data capabilities. So, um, you know, we've, we've gradually increased the data volume you can put into Airtable from like tens of thousands to hundreds of thousands. And now we're about to launch like hundreds of millions. record database engine. But these are some of the things that we had to kind of, you know, it's not just how you sell and like, you know, the way you go in and target buyers. It's also, it has to propagate all the way back to like, do you have a sellable value prop?

And that speaks to the critical relationship between sales and product.

For sure.

Within an organization, often that relationship is quite dysfunctional.

Very.

But having a really integrated relationship is key. You say, Howie, that it's important to have an almost surgical understanding of your company and really understand the key metrics that drive the business. And, you know, Villages, and Duane likes to talk a lot about the importance of instrumenting your business and really understanding the levers. Talk about what you've learned as a CEO on this journey. What metrics do you track today? What metrics were you tracking at the beginning? How did you determine what the key levers were?

Yeah, I would say it's been a constant journey, and I don't think at any point I don't think I've ever felt like, oh, we figured it out. It's like perfect, right? And I think, I mean, there's frameworks, like obviously you can do OKRs, like there's Salesforce V2BOMs, like Amazon input metrics. And I think we've tried some version of all of these.

What is Amazon's input metrics?

So the idea is basically instead of just tracking against output metrics, which would be like the revenue of our business, input metrics are the basically levers that you believe most drive the output, right?

Yeah.

So one input might be for them, like, okay, how many, like, what's the engagement, you know, on this specific type of item page for this category? Like, how long do people spend, like, looking at the product on the books page as a kind of leading metric that they want to focus their execution on? Because, you know, it's hard if you just say, like, okay, grow revenue. Well, like, it's hard to know what to do about it. Whereas if you say, like, okay, you gotta increase engagement on the listing page, then you can say, okay, well, I wanna build like a more interactive experience to like page through different photos of the book or see a preview or that kind of thing.

Makes sense. By the way, just on V2MOM, do people know V2MOM from Salesforce? It is kind of amazing. I feel like 20 years ago Marc Benioff came up with something that sounded nice to say, like V2MOM is kind of a fun phrase to say. And then it remains the governing framework inside that entire organization and now it's been emulated. step back and think like just some like 20+ years ago he came up with this. It kind of worked at that scale. I'm not sure like companies around the world should emulate that. So it is funny sometimes these frameworks, like Salesforce has been very, very successful. I'm not sure it's because of E2MOM. And there are probably other better frameworks for organizing a company.

Well, here's actually my punchline is I actually think the framework does not matter. Like you need some kind of framework. What's your framework? Well, so what I would say is like, and this is inspired by like, you know, Atomic Habits and the whole idea of like, identity formation actually matters more than like framework. Like framework in a health context would be like, okay, I'm gonna have a routine, means I'm gonna go to the gym these days and like not eat these things and blah, blah, blah. Like, but it turns out you could do that, but adherence to the framework, especially when like, you know, exceptions arise, you're traveling, whatever, is like, is where it all breaks down, right?

Mm-hmm.

And I mean, similar to like, you know, let's say for product development, Job to be done is a great framework. It can be a great framework, but I've seen it applied really poorly, like at many companies where it's like you literally check the boxes on like, what's the job to be done for this customer? Like, who is the protagonist? You know, what are they like? And so my point is like, these frameworks can all be applied in a really poor way if you don't fill it in with like really the right, you know, framing of, you know, the initiatives you're pursuing, why you're doing them. And then more importantly, I think having the urgency and accountability for follow-through, right? And I don't think that gets solved with just a framework. Like, you can do RACI and say, like, okay, here's the responsible party and the accountable party and, like, the, like, you know, informed and so on. But I found, like, and, you know, we've tried all these things. And I think what I've found is much more impactful consistently is having leaders and leaders, I mean, not just at the top level, like exec team, but at any level, who, you know, just really are motivated to figure out what matters and are able to kind of navigate ambiguity. I love this book, Extreme Ownership, by like these SEAL team captains. Right. And you think about like that kind of context, it's life and death, high urgency, and battlefield conditions are changing dynamically. Right. Like things are never how you think they're going to be when you go into a battlefield situation. And so that, that metaphor is really powerful because it's more about the embodiment of you know, kind of leadership where you're able to make game-time decisions in the face of constantly being curious and learning and like, you know, gathering intel.

Yeah, well, and again, I wanna come back to the metrics thing, but I think you're making a really important point, which is following through on stated values or frameworks matters a lot. So to defend V2MOM, like the genius of the V2MOM framing is it's memorable, right? Like here we are talking about it.

For sure.

And just if every, 'cause the ring of the phrase resonates, it's like Tom Friedman, coining the world is flat after people have been talking about globalization for hundreds of years. But hey, the world is flat. That sticks. So like coming up— that's why I think names of companies and products are deeply underrated. The name of your company, the name of your product. I think Airtable is a great name. I think Village is a great name for a venture fund. I think these names do really matter. Entrepreneurs should spend more time—

I think of Yellow all the time. It's like glued in my mind.

Howie was at our gathering where he was supposed to do a one-on-one brain date with Mira Murati, The day before she was announced, or she was getting the news that day that she was becoming CEO of Twitter.

Yeah, I think she was literally probably pulled into some emergency briefing meeting at that point.

Yeah, you took it personally and then realized she had beat up—

I wasn't personally, yeah.

Yeah, better reason. But okay, so coming back to Airtable, what are the key metrics that you measure today?

So, you know, at the very top level, starting from output, obviously we care about revenue, we care about NDR. NDR actually, I think, is one of the most important metrics, especially once you've gotten past the initial product market fit stage, like the difference between a company that can continue to compound at a very high rate, like take ServiceNow, I mean, it's still growing at an incredible rate.

Yeah.

$10 billion in revenue. And part of that is because they have really high NDR. Like they can just take every customer cohort and keep growing it. And so for us, like tracking our NDR is actually one of the most important things. Now obviously that's again, like very like lagging. You know, we look at the end-to-end funnel in a way that, you know, tracks like the top of funnel web traffic. Like we look at the number of like what we call multi-user active. Workspaces. So in the self-serve, like top of growth engine, like how many of these email domains or users have come in and then activated and then added at least one collab. So it's a multi-user use case. And of course, like tracking that through the funnel, we track against specific initiatives that we're really kind of trying to make progress on. So AI is a big one for us. So we're looking a lot at like the AI usage, where customers were most active on AI. I tend to think it's not like sometimes in the 0 to 1, and not just like a company going 0 to 1, but even for a new initiative, like trying to break through on finding product market fit for a new capability of ours, right? Like if we're launching a solution on top of the platform, or if we're kind of launching these new AI capabilities, that's a 0 to 1. initiative. And I think for that, you care more, you care about more than just like the top line metric, like how much AI revenue are we generating, how many customers, et cetera. You also care about like kind of debugging and understanding like who are these customers, what are their use cases? It's much more like the, the Lean Startup, like kind of really rapidly learn and figure out like what are the hotspots there? And then how do you kind of lean into it and fan the flame? So for instance, I just set up an auto emailer today That's going to take batched sets of all of our top, uh, topmost active AI customers and basically email them from my actual email account. And I'll actually get the response asking them, what are you doing with Airtable and what's your use case? And, um, I actually did this back way in the day when we founded the company and when we launched for the first time. But that kind of qualitative insight, um, I think is as valuable for the zero-to-one kind of breakthrough initiatives. as the overall KPIs, right?

Yeah, it's the Jeff Bezos line. It's always day one. And even in a big organization like Airtable, there are all these brand new zero-to-one initiatives. Let's talk about AI. We have in the room today, I just saw Guy from Perspective AI come in. Where's Guy? Do you want to just give him the mic? Do you want to share what you're up to? Because we have a couple vertical AI companies just to get a flavor, and then we'll come back to Airtable. Do you want to introduce yourself? Sure.

Guy Nirpaz, big fan of Airtable.

Thank you for doing this. Perspective AI is Basically the qualitative feedback. We are— the idea is to get feedback, you have surveys or conversations.

So surveys, you get the scale, conversations where you get the depth.

And we're using AI to have the scale of conversations.

So scalable quality conversation, that's the proposition.

We're kind of relatively early. Cool. Great. Starting with PLG. Perfect. And then Dion, do you just want to talk about AI and customer support? We'll hand you the mic. Hi. Good to see you. So we are Forethought. We build the most advanced generative AI agent for customer support. We're a fully agentic model, help power businesses like Airtable, Asana, Lime, many others.

I think customer support is going to be one of the most impactful use cases of GenAI. And so excited to hear your thoughts on that.

Yeah. Awesome. And there's hundreds of others, of course, some others in this room and all around the industry that are being AI-first startups being funded today. And many of them have companies like Airtable on their sites, not the entire business, but different verticals that they could attack you on. How do you think about that competitive dynamic? And if you were— because you can imagine starting Airtable today in this AI moment, it probably would've been a pretty different zero-to-one launch story, right? How do you think about that?

Yeah. Well, yeah, I mean, I don't think we would have taken 2 and a half years to build the product. No, I think we did build a lot of the foundational architecture of Airtable, and that has continued to pay off as an advantage. I just think if you're coming into the space now, you couldn't afford to do the same. You'd have to take a different route to building the MVP, which actually I think is what's happening. So I think you're seeing a lot of products, AI products that are taking a much more agentic approach to begin with, which in some ways has, you know, like a different and maybe simpler, like, zero-to-one, you know, kind of product building challenge than, like, trying to build, like, a completely new, like, database engine and then creating, like, you know, the workflow layer and, like, the UI, et cetera, on top of that. Um, you know, and I think there's pros and cons. I think a lot of the agentic products will do well. I mean, some of them are already taking off, like developer copilots.

Um, Mm-hmm.

Like, there's a lot of, you know, kind of revenue growth happening out there. Um, and, and, you know, a lot of, like, market TAM to be seized. For us though, we're going to lean into our foundational advantage and we're going to use that to kind of maintain an edge and actually to gain more value from AI by deploying it into data and workflows that we have, right? So the idea is in Airtable, you're already putting things like your product roadmap or content production pipeline at media companies, or it could be like all your marketing campaigns at a really large retail company. And since you already have that in there, and it's very, you know, it's a very good, you know, structure for that data and the workflows. And it's got, you know, kind of a wide footprint of people and using that, that, that use case, it's very easy for us to introduce AI into that workflow, right? So the data is already well structured. You can then take that, put that into a prompt, and then generate some output.

And how do you think about human in the loop versus fully automated?

So I think for some use cases, which are not the ones that we serve, you can do fully automated, right? Like, I think support will be one of the ones where, especially for simpler, more linear support workflows. I mean, look at DoorDash, right? Like, I don't even need to talk to AI, let alone a human, when I go and, you know, ask for help on DoorDash because it's very simple in terms of the decision tree, right? So it's like my order came late, like, okay, hit this button, hit that button, auto refund, right? It's just like it tells you, it does it automatically for you. It's just a simple rule-based system. And I think that works. And so therefore you can imagine, like, with a little bit more reasoning capability, You can actually automate slightly more advanced support workflows quite easily. So I think there's certain verticals like that where it's very linear. It is something that's fully end-to-end solvable with AI that will be solved. I think for most forms of knowledge work, there's too much nonlinearity and too much kind of fuzziness in the system. And that is where the human steering, the human in the loop, the human not just to perform a task for the AI, but vice versa, where the human's kind of orchestrating the whole thing. The human's like checking the outputs of the AI, especially when you have like, you know, high sensitivity, high stakes use cases where you don't just want the AI to like, you know, go like in this recursive loop to, to like, you know, the complete solution. You want to have checks along the way. I think that's where, you know, AI I see as really much more of a, you know, a really powerful, um, you know, function call that you can embed into different parts of the workflow. But you absolutely need the data and the human in the loop aspect of it for it to fully solve the problem.

We have a lot of companies here selling to enterprise with an AI-centric product. Do you think there's enterprise fatigue around the AI promise? So much hype the last year or 2. Some would predict that the next 12 to 18 months is going to be this kind of big step back where companies say, okay, we scratched our itch of curiosity. Where really is value showing up? And actually, it's sometimes astonishing when you double-click a little bit in these bigger companies how little AI they're actually using, right? Totally. There's press releases.

It's like the water, water all around me, like not a drop, right?

Indeed. Yeah. So where is the use case today in enterprise where AI is actually having an impact and where do you see it going next year?

So I completely agree with that premise. I mean, we see it in a lot of our enterprise customers ourselves where, you know, like there's a backlog of, you know, it used to be like a year and a half from now, it was like, okay, like we just want to buy every AI product that we can because there's not that many offering it to us, you know, that are coming to enterprise, right? There just weren't that many products. that were going to enterprise a year and a half ago. And now, like, that space has just been so explosively inundated. So I think the enterprises are saying, like, whoa, like, you know, now before we just wanted to, like, go and experiment and learn from, like, the early couple products that were, you know, were available to us. Now we're getting pitched by every existing vendor with their AI add-on, whether it's Salesforce or Slack or whatever, you know, or like there's a net new AI vendor that's trying to pitch us on their thing. And I think there has been also a lot of hype and promise around, like, this is the AI solution that will solve everything, right? You just put it across your entire org and it will solve all your problems of AI. And I think the reality is, of course, these companies are not seeing massive kind of business outcomes, right? They're not seeing double-digit percents or even basis points of cost reduction across the org as a result of these AI deployments. And so it's kind of like, okay, you know, when the rubber meets the road, we're not seeing the results. So I do think there is, there is a reckoning that's already happening now. And—

So what advice would you give to early-stage startups when confronted with that reckoning with selling into enterprise?

I think, I mean, we're doing the same, which is we are selling against specific use cases. And in our case, because we're a platform, it doesn't have to be just one, but we're going in and selling against specific use cases where we have proven value both as a platform period, like for that use case, but also where the AI has a clear role to play, right? So as an example, like one of the use cases we sell to is product operations. So if you have a digital product group within a large company, it could be like a large retail company, large tech company, whatever. You have these large software product groups, maybe you're building the e-commerce platform for your company. And first of all, we win that use case frequently because Airtable is a good flexible platform for that use case, like Salesforce is flexible for CRM. But then on top of that, we'll come in with an AI point of view of saying, look, what if you could take all of your customer feedback from different support channels, from app review sites, from so on, and ingest it all with AI and basically come up with insights? You can triage certain feature requests and like, here's the specific customer, maybe tagged via Gong call that is requesting this kind of feature or wants you to do this kind of thing. And so that's a very specific application where we can actually go and say, like, here's the value you're going to get and you're going to get it.

Look, and I think you're making an age-old point in a sense, which is your advice is focus on what the customer problem actually is and solve the customer problem, which sometimes I think in this AI hype cycle, founders can lose sight of that. It's like AI whiz-bang, you should buy it versus obsession with understanding the customer need and articulating it. Yeah.

And I think it has to do with, like, I mean, maybe we falsely thought, or there was this feeling of, like, we're in the tornado of AI adoption, but I think we're actually in the chasm. And so this hype was actually just like, to your point, it was relatively ungrounded in real business value. Now, at some point, we may get to the point where it's well understood what are the categories of AI value exist, and then you just buy the obvious.

How long is the chasm going to be if you had to predict?

Well, it's hard to predict because I think unlike most other technological disruptions, this one, the technology itself is progressing so fast. Right. It's like one thing when mobile came out and it's like, okay, what are the mobile-first applications we need to deploy into the enterprise? Well, mobile didn't change that. The form factor and the affordances enabled didn't really change that much after the initial innovation. Whereas obviously we're dealing with an AI progression curve that is actually unlocking dramatically different capabilities at every point. Right. And it's moving faster than even Moore's Law. So it's like you're trying to predict exponential math with a really unknown coefficient.

By the way, these are the very circumstances that might give an edge to startups that are lean enough and have low burn and low headcount to navigate as the new tech unfolds.

I think there is kind of a race between incumbents who obviously have a distribution and sometimes data advantage, which are really valuable in this world of AI, But they do have to innovate.

Are you an incumbent in this?

I would say we're kind of in between. We're kind of an adolescent teen. We've maybe barely gotten past the pimply phase, but we're basically somewhere in between. Right. And my point is, I think the question will be, will the incumbents innovate in a real way fast enough and not just do token things with AI? Right. Not just put AI on the website or say, here's our AI offering, but it's not actually driving real value. faster than the startups can go and gain significant enough market share and build enough of a durable advantage to withstand when the big giants come down at them. And I think we're kind of somewhere in between, which is we're obviously able to be a lot more agile than the big incumbents. We do have some more distribution advantage than a net new startup, but we also have to still move quickly, right?

Yeah.

We don't want to be caught Yeah.

So let's shift, and then I'm going to ask a few more questions, and we'll open it up for your comments and questions, everyone.

Yeah.

But let's talk about advice. One of the most important things a founder has to decide is who to take advice from on what types of topics. And it's usually the sign of an inexperienced founder when they indiscriminately absorb whatever advice gets thrown at them. And I think as investors and as advisors in the ecosystem, do no harm has to be the first That's the first rule that we follow, and you can do a lot of harm with bad advice.

For sure.

When you reflect on the last decade-plus of Airtable, what's some bad advice that you were given by investors, board members, advisors, friends, whatever?

I mean, I think all advice is good in a different circumstance, probably. Maybe not all, but a lot of advice is good with a different set of circumstances. But I think it's incumbent on you as the recipient to kind of understand what the priors are of the person giving it to you, right? So like, you know, we got a lot of advice early on to like not try to build a platform company and to just pivot to like one solution, right? Including from really great founders and operators like Marc Benioff. Like when I pitched him on Airtable, he was basically like, I think you should go fully into like healthcare, do like some like EMR thing. Like there's a very clear need for that.

That is the conventional advice. I mean, I think focus on a niche is the common advice.

Yeah. And by the way, he probably wasn't wrong. I mean, look at Veeva. Like that was a success. Yeah, yeah, incredible. Yeah, not like a bad piece of advice, but clearly we were able to find a different way building more of a platform business. Now we ended up having to do some solutiony things on top of it early on with templates and different PLG entry points, like people coming in Googling for one use case or another and then landing on that template or landing on a marketing page for that use case. And even now, we are doing a lot more solution selling, even if the underlying platform is, you know, kind of the same once we deploy or, you know, when you start the deployment. But, you know, that was a piece of advice that was right given a different set of circumstances. Like, it probably would have been the necessary path if we were enterprise sales focused, because you can't enterprise sell a Lego kit that's all things to all people.

And how do you do that just as a founder? You know, it's basically your first time founding. You had a company sold, but you're pitching to Marc Benioff, legendary CEO. Why is Marc Benioff coming up so much in this conversation?

I mean, I worked there. I have respect for them. They're a $30 billion revenue business.

Yeah. So here's Marc telling you, focus on a niche. Every blog post about startups would say, focus, dominate a niche. And then you say, nah, I'm going to build a— I'm going to spend 2 and a half years building a horizontal platform. How do you develop conviction to go against the grain like that?

I think there were a couple of products that had done something similar, like Dropbox. Um, you know, obviously it was a very horizontal product, but I think they were able to break past the conventional wisdom because it was such a stupidly easy product to use, right? I think if they had much more complexity to how you used it, they would've failed as a horizontal product. They would've had to go like down into some deep vertical, you know, use case like DAM where you could sell a lot more premium seats or, you know, premium solution, but, you know, it's more complex sales and deployment. But I think the fact that it was just so drop-dead easy to use meant that even without a crisp understanding of what are the specific use cases each person is going to use it for, like, you could use it for photos or document editing or kind of this team-centric use case of, like, let's put all the resumes for recruiting into this folder and collaborate on that. But the fact that it was so intuitive and you could just show the thing, share the actual file, share the actual folder, and the second you came into it, you just got it. like allowed them to overcome, I think, what traditionally would have been like, you know, like marketing too broadly to too many people.

So you found a couple of examples that provided inspiration to say, look, it is possible to blaze a different trail. Yeah.

And I think, you know, I also think like we tried to do, you know, a platform with some, you know, kind of use case marketing on top, right? Like the templates. So I think in a way we tried to do both. I actually did read this book at the time called Relationship Marketing by Regis McKenna. He was like an—

Yeah.

You know, kind of marketer. And, uh, I mean, obviously very different time and like, you know, so on. But, uh, I was, um, you know, inspired by the fact that like they talked about marketing for the Macintosh, which I guess, you know, it's worth noting it wasn't a very big commercial success. So, you know, but, uh, what do you wanna extrapolate from that? But, you know, but they, they thought about it really as like, this is a platform that is gonna be so like broadly applicable. Like think about all, I mean, you know, it's the first breakthrough personal computer that like anyone can actually use, right?

Mm-hmm.

Like everything before it was like, complicated, more like, you know, like the Apple II required like more of a, you know, real nerd to, uh, to get to use. And so this was like the first mainstream computer, but like, how do you market it? How do you sell it when like there's not, you know, you could verticalize it and say like, we're just gonna sell it for like engineering applications or for like graphics design applications. But of course, that would be like a real shame, right?

Yeah.

The whole point of it is like, it's this broad computing paradigm. So I think in that case, it's like you can still think about the different personas and use cases, And even try to weave those into the storytelling of like, you know, when they marketed that product, they actually did talk about specific use cases like, hey, you can do graphics design. Look at this beautiful, like, paint, you know, painted thing you did.

Yeah.

So I think we tried to do some of that as well, which is we didn't just talk about the platform in the abstract. We, you know, wove in examples of use cases without committing to being just a vertical solution for that.

And I think just in a meta thing on advice, like If you can point— if you can find one example, like, story from my own life is that when I was thinking about stopping my formal education journey, I reached out to like several dozen people and said, hey, should I just like not do college and just do real-life university? And every single person said, no, no, college is really important. And one person, Brad Feld, the VC at Foundry Group, said, oh no, it's a total waste of time, you should just go do your own thing. And I clung on to that one, because that was my intuition.

Yeah, confirmation bias maybe.

Yeah, exactly. Because we all have instincts.

Yeah.

Right? And you had an instinct as a founder that this was the way to go, and you needed to find— like Dropbox is a great kind of example to drive you in that way. What bad advice on hiring, raising from VCs, things like that, as you reflect on the journey, what other things, mistakes you made, or people told you, oh, you gotta pitch your company this way, or you gotta hire this way, that you—

Well, by the way, on that point, like, I do think, you know, I think founding a company is sort of like being an investor, where like if you do everything else that, you know, if you just do the consensus thing, you're not going to have alpha. You're not going to, like, you know, generate, like, a breakout return, or in this case, like, a breakout innovation and, like, capture market in a profound way. Right. So, like, Peter Thiel had the concept of, like, secrets in his whole, like, Stanford class lecture series. And I really liked that idea of, like, okay, like, you can't just, like, reinvent the wheel on every single thing, but you can have, like, a few key secrets where, like, you have deep conviction that that hopefully is in some way empirically grounded. So it could be market evidence, or you worked in the industry, or you have some unique take. It's not just conjecture, but you really feel high confidence, high alpha, that there's some misunderstanding or mispricing in the market around this thing, right? And you kind of go all in on that. And of course, you can validate that. I mean, your thing was clearly like, don't go to college, right?

Yeah.

And I mean, most people would tell you that's wrong. But you clearly— I don't know how you developed that initial instinct, but you had some kind of alpha, and you got just enough confirmation of it from the right person to where you're like, OK, I feel confident about making this bet. I'm going to hit buy, right?

Yeah, and I think this idea— I can almost imagine the slide. We think about this at Village raising funds from LPs, but I think companies are the same, which is like, here are the 5 consensus views that we hold, and then here's the one non-consensus view we hold really vigorously.

For sure.

But you can't reinvent the wheel on every dimension, right? Right. And generally, the way everyone does something or the life path that everyone takes generally is the right path. As Bezos says about contrarians, most contrarians are wrong.

Right.

Yes. Markets are efficient.

It's like if everybody jumped off the cliff, well, like, yeah, I mean, maybe there's like something to it, but like, you don't want to just go jumping off cliffs on your own, right?

Indeed. But like the one area where you are, the one or two areas where you're truly different.

For sure.

Really. But I think when you articulate this to investors or to customers so that you don't seem like a crazy person or it's kind of reflexively contrarian because that's the high status thing these days, articulating like, no, no, we actually are consensus.

When I was growing up, it was called emo.

Yeah, exactly. Yeah, you want to be consensus and signaling that you actually are consensus on most things, like, gains— gives you credibility to make the contrarian argument.

Exactly.

Yeah, I think that's right. You don't want to come in and just be like the completely alien spaceship to the car that the customer is looking for. At the same time, you don't want to just You know, you want to come in with like a different point of view in the right ways, right? For Salesforce, it was cloud, right? It was like, we're going to come in and we're going to tell the story of our belief system, which is all about like cloud software and the cloud model, which is a superior way to build and deploy software. And you're going to get a better experience because of it. And in fact, all of this customizability would not be possible if it wasn't for this cloud, you know, kind of model that we've taken, which at the time was pretty radical, right? Like, I mean—

So we're going to open it to questions. This is my last question for now, and then we'll open it up to you all. Howie, you're about to have a kid. Thank you, yeah. You were raised by tiger parents.

Yeah, more or less.

Your own description. What were the— and you've said that the tiger parenting that you received was net positive for your own maturation. What were the pros and cons? And when you think about raising your kid, are you gonna take a similar framework and like, you know, drive them hard?

Yeah, so first of all, I'm not gonna take the same framework. You know, I think like, I mean, similar to the advice and like, you know, the practices like, right thing for the right time, right? I think, you know, what, um, the parenting style that my parents, especially, especially my mom, had was both born of, you know, kind of a survival mentality, you know, like, um, a third generation, whereas my grandparents fled, uh, during World War II from Korea to China and like kind of, you know, went through a lot of hardship during that time. Like, you know, um, like there was near starvation at times, right?

Mm-hmm.

Like there was like real, like like literally life or death survival. And of course, like growing up, you know, through, through that era, you know, Cultural Revolution, my parents like grew up in, you know, kind of modest, you know, circumstances. Like I grew up in modest, very modest circumstances. So a lot of it was like—

Wait, where did you grow up?

In Texas.

Okay.

But, you know, but like—

We got some Texas. Anyone else from Texas here?

All right.

Yeah. Subscribe. Well, there'll be a meeting at the back. All right, great.

Yeah. Texan alums. But, you know, it's like a lot of that was like, hey, like, you know, you can't afford to go and like just like wander around to figure out like what you want to do. Like, you have to like be driven and like have a clear directive and purpose. And so I think that was the right like mandate, the mission, you know, given the circumstances, right?

When did it take your parents to realize that Airtable was a successful company, or do they still think you should be a doctor?

No, no, no.

Like, my mom especially always kind of like understood, like, there's something to this whole tech and software thing. And like, you know, like, I think she was always supportive of us. Yeah, supportive for sure.

Yeah.

Like, I mean, actually, like graduating from college. So first of all, I thought looking at like all the, the, you know, the statistics on like great founders like Bill Gates, Steve Jobs, like Zuck, like they all dropped out of college. Right. And so I was like certain that like graduating college would be horribly anti-correlated to like, you know, startup success. Founder success. And so I was like really bummed to actually graduate from college and not have dropped out. Like, I didn't find like a good idea that was worth pursuing. But when I did graduate and like I eventually had my idea for eTax and like decided I wanted to go down that path and like give up actually a job that I already, you know, already had signed and kind of, you know, had lined up that would have been like, you know, great job paying well, like blah, blah, blah. But getting that offer.

And then not accepting the offer is kind of like dropping out of college in the sense of you get the status.

Oh, worse. I actually, I had the offer. I don't recommend this and certainly don't recommend this if you're joining Airtable, but, but yeah, so I had accepted this offer and like literally like months after accepting it and like weeks, like, well, actually the day of the start date is when I had this pivotal moment where I really felt torn. I was like, should I make the leap and like, you know, do my own thing? Should I just take this job at least for a while? And I mean, I literally, it was such a big decision for me because I also didn't have a fallback. I didn't want to go back to my parents and need to rely on them because I knew they didn't have the luxury of being able to keep providing for me.

Right.

So it was a really hard decision, and I basically delayed it almost by accident because I couldn't make a call until the day of. Time ticked by, like, you know, I was supposed to be there at 8 AM for orientation. It's like 8 AM, I'm like thinking, like, do I want to go?

Do I want to—

you know, at 9 AM, like 10 AM, I'm like, okay, well, I guess I'm like— I think I got a recruiter call. They were like, are you okay? Are you not gonna— like, are you, you know, on your way? Are you late?

Got in traffic.

Yeah, and, you know, in that moment, I just decided it was like the, the impetus I needed to make that call. And I was like, you know, I'm really, really sorry. I know this is like gonna cause some distress for you and You know, it's day of, it's like kind of bad form on my side, took accountability, but like, I'm really sorry.

We're not gonna name the company, but it does rhyme with Google.

No, but like, yeah, I basically decided like—

But I think there's actually a really interesting insight there, which is why so many big deals and negotiations fall apart at the 11th hour, because people wait until the very end of a negotiation to say what they actually think, want, or believe. And so like, it's an interesting question to make better decisions, can we kind of Reinforce that in ourselves because that's how we can really tune into what we actually want.

For sure. No, I think that's right. And I mean, there is a fine balance between, you know, being too impulsive and just like, you know, quitting your job on a whim before you have like the right validation, before you have the right like alpha to go and like, you know, start a company with conviction. But on the flip side, like you don't want to wait and keep delaying it because there's never a perfect time, right? So I think, yeah, it's kind of finding the in-between. I find that as now an operator and CEO at scale, the best way I would translate that into my practice of leadership is trying to create urgency around everything that we do. If there's ever a moment or a day or a meeting or a project that I'm working on that doesn't feel urgent, I try to say, why are we doing this? Either we don't understand the real thing that we're trying to figure out, or we shouldn't be doing it. And I think that's a way of sort of like forcing yourself to make sometimes the hard calls. And it could be like, don't keep doing this, it's not worth it. Or it could be like, you gotta pivot what you're doing. Like, and instead of delaying that by 6 weeks because like it's the easier thing to not hurt people's feelings and say like, cut what you're doing and like, you know, whatever, but like, do it now.

I think the mistake people make emotionally on decision-making is they, they, they are waiting for themselves to feel ready. You just said there's never a good time. Yeah. And sort of a frame that I once thought of on this, on the readiness concept, was like, do things when you feel almost ready. And I heard, I heard this from someone who does a lot of skydiving, and they said when you're a beginner skydiver and you go up on the plane, then for your very first dive, the instructor will say, okay, on the count of 5, we're gonna jump.

Yeah.

And what happens is most people, when you get to like 3 or 4, they say, no, no, no, no, no, no, no, don't push me out.

Yeah.

And so you go 1, 2, and then they push you out on 3, right?

Good to know.

And, and so like You'll never feel ready to jump out of the plane. And so the litmus test— and like, you'll never feel ready to start a company or to ship this product or to make a bold decision. And if you keep waiting for peace and readiness, you'll wait too long. So like, almost ready is like the internal emotional barometer for action. For sure.

Yeah, I think, I mean, there's a fair amount of discomfort that is just inherent in the founding and operating journey. Like, I found at least, I mean, somebody else has had a very easygoing time of it, like, let me know how to do it. But, but yeah, I mean, like, I think I've almost kind of just become tolerant of it almost like, you know, not that it's like masochistic, but like, you know, you almost like start to enjoy the discomfort in a way because like you've learned the right amount of discomfort to push yourself to, right? Like, you know, the— like, I remember like, you know, in that 2 and a half years we spent building the product, like, I was always uneasy about like, oh, but like, it's still not ready. We could spend more time building this. Just another month. Like, we need to build this one more thing. And, you know, that, like, that's probably the eternal thing with all MVPs. It's like, how, like, at what point do you draw the line? And so, you know, I think we kind of had to force ourselves at some point to draw the line and, like, push, you know, launch with discomfort. And I'm so glad we did because, like, if we had waited until we got comfortable, it could have been like 5 years later, right? Yeah.

So, so, like, wishing, you know, having aversion around discomfort or fear or whatever is the, is the path to unhappiness.

Yeah.

Figuring out how to, you know, but, but let's come back to your kids. Um, 'cause, 'cause this is the thing that like every, every person I know who's like somewhat successful, they have kids and they usually had a hard upbringing themselves.

Yeah.

And the thing they're worried about is how am I gonna make sure my kids work hard and like have to hustle and have some of the values that I grew up with?

Yeah.

Um, so how are you gonna solve that?

Well, I don't actually need my kids or care about my kids like having the same level of like hustle that I do. Um, 'cause I think there's different like utility functions in life, right? Like you can define fulfillment, happiness, purpose, like, you know, kind of life goals in different ways. And I think, again, like, for me, that intensity and grit was like kind of a necessary mission because of like the circumstances. But, you know, like, I, you know, I feel fortunate that like hopefully my kids won't have to like have the same amount of grit because like, you know, they can have a little bit more buffer and leeway to like go and explore and like be, you know, if they want to do creative pursuits, like, great. Like, you know. So, you know, I would encourage them to like find—

Major in sociology and write poetry.

Yeah, like, great. Well, I mean, really prompt engineer AI to write poetry.

That's right. Yeah, not poetry. That's endangered. All right, let's go to questions. Raise your hand if you have a question for Howie, and we'll start with Rishi.

Yeah.

Hey, Howie, thanks so much. Rishi Toparia, co-founder of Groot Ventures, pre-seed. You know, going back to the topic of the move from PLG to the enterprise.

Yeah.

You know, when it's the PLG motion, people are looking for the product, they're searching for something, landing on a page.

Yeah.

Almost seeing that kind of instant day zero, day one delight. As you've now migrated to the enterprise motion, to the point on enterprise fatigue with AI, how have you taken some lessons from the PLG motion on that early delight for the customer and applied that to your enterprise sales process? Yeah.

So I think of it as like, how do you create credibility with the customer? Right? Because they're— I mean, even pre-AI, they were getting pitched by a million different vendors. Right. And all of them are making promises. 99% of which are probably never going to pan out. Right. I remember talking to a CFO who basically said, if every company that claimed to save me X dollars or Y amount of hours of time actually promised on that, I would have like employees working -50 hours per week and driving 5x more revenue, which clearly was not the case. So, you know, I think like there's already been like this inundation of like false promises. And so I think it's more about like a credibility question. Right. And it applies also to AI. And I think there's multiple ways to build cred. One is if you already have traction or proof points within the product or within the company. So in our case, the PLG actually directly helps because we can say not just, hey, you have this much usage, which is not value in and of itself, that's just usage. But if we have actually done our research and understood what are the useful outcomes of that PLG adoption. So if there's a user research team at XYZ really large tech company that's already built an awesome use case around sharing and extracting insights out of research notes, and that's being broadly consumed by a lot of product groups that are saying, hey, we're actually able to build products better with that. That's one form of credibility that we're going to include in our story. I think a second part is case studies, obviously being able to say we've done this for other customers. The third is platform and architectural differentiation. be able to actually say technically and tangibly, we can solve this problem that you have better and different from other people. That's another form of cred. So I think it's really about like, you know, knowing that you're not the only one making a lot of promises to the company, to the customer. Like, how do you differentiate, you know, your ability to actually deliver on that?

Yeah.

And there's—

and I'll just— and I just to reiterate, because it's like we talked about this a lot at Village, which is like experts understand simple things deeper. Differentiation. That in and of itself is not like headline news, but few founders internalize that as deeply as they should. It's like everything we know, the advice for business, but it's how deeply can you actually act on it?

For sure.

Like obsessive about differentiation and being able to articulate is key.

Yeah. I also think like we, you know, in the PLG era, I think we as an industry became very like navel-gazy, as in like, you know, you build this product, it's so beautiful, it's got amazing UX, people love it. And you measure NPS and it makes you feel even better about yourself. Like, oh my God, look at how many people like think that, you know, they would promote like this product. But in actuality, what I think it can fail you on or what you fail on is like you don't actually understand like what problems you solve for the customer, right? Because, you know, you orient— you start orienting yourself around like the features and like what you like about your own product, right? Like, what am I proud about my own product, right? Versus do I really understand What the customer cares about, because the differentiation shouldn't be self-oriented. Like oh, my product has better colors than the next one or better UX. It's like okay, but how does that solve what value we deliver?

We had Lucas Bywell at our at our event recently from Wait to Buy. He said however many customers you're talking to, you're not talking to enough. Yeah, like that's just like you've got to be talking to more customers. Sure. Let's do one or two more questions. Yeah, Arnel.

Arnel, one of the founders of Edlift. We love Airtable. Can you do a deep click on how you're design partnering?

Customers, design partners.

Yeah, so early customers and balanced what they wanted with your vision of what the middle are going to be.

Yeah, so I'll try to answer with a short version, which is I think it's a little bit easier when the customer is fairly small or the use case is smaller. So small teams, individual, et cetera, because I found that the larger the enterprise problem we're trying to solve, the more hairy it is and the more temptation there is to go and like say, okay, but like to really fully solve it, we need to do all these things. At which point, like it's a prioritization challenge and really like distilling the essence of the problem down into not what the customer thinks they need because they have all these requirements that actually may not be true requirements, but like sometimes helping them unpacked, like, what is the actual problem you're trying to solve, right? You don't actually need this, you know, specific form of data validation rule or this form of, like, approval checklist if the real thing you're trying to do is X. That's something to reflect on.

That's a deep point. The idea that the customer might not be able to actually articulate what they want and you have to help them is kind of profound.

I think customers often try to jump to the solution. They tell you, like, hey, here's the feature, here's what I want it to look like. And what you need to do is to actually try to unpack underlying that. What is the real job to be done? Because otherwise, if you become just a feature checklist kind of machine, you're never going to have a breakthrough product and you're going to be torn between this customer wants these features, which are fundamentally in conflict with this set. And so it's always like, how do you go unpack the underlying problem? Now, I found it was easier to resist that temptation to just build features and kind of allow the customer to solution for you and for you to become an order taker with the smaller teams. Because I think you just come in, you don't expect a 5-person team to tell you exactly what to do. I think you go in inherently with more of a mindset of like, okay, I've built a workflow tool for them, let me show them how to use it, and then let's see where they get stuck. Or like, oh, it turns out to really fully solve the problem, they have to have this interface type, or You know, kind of this automation or integration, et cetera. But I guess, yeah, maybe to simplify, like, understanding what is the real problem to be solved.

And then, um, yeah, and if a customer starts telling you what product to build, the conversation's gone off the rails. And a customer development conversation should be about their problem and experience, right? If they start trying to tell you what features to build, it's like they've lost focus, which is a really profound point.

And if you can't get them to tell you the, like, if they can't unpack like the underlying problem to be solved, then maybe they're not the right customer to get that feedback from. You still want them as a paid customer, but I mean, same as your point about knowing not to take advice from too many people, customers are the same. And I think we've often found the ideal is a customer, a specific person who is kind of visionary and strategic in a way where they understand their job is not just the clicking of the button. They understand, okay, let's say I'm in HR, my job is not to go and click buttons in Workday. It's actually to most efficiently manage, let's say, payroll and time off or whatever it is, right? And if that's the real problem to be solved, you can actually go past the feature requests and into, okay, what's a better overall approach to this? Maybe you can have a chatbot agent that helps you do those things, even though the person's gonna go to, I want a faster horse. I want like a button that looks bigger and is easier to click.

Awesome. We're gonna leave it there. Howie, we really are grateful that you're part of the Village, and I appreciate you doing this for our founders and some great insights. Thank you so much.

Yeah, well, I appreciate being part of the community.

Yeah, thanks so much for listening to the Village Global Podcast. You can check us out online at villageglobal.vc. We'd love to hear from you, your feedback, your ideas, your inspirations. You can email us at hello@villageglobal.vc. Hey everybody, this is Ben Kesnoka, co-founder and partner at Village Global, a network-driven venture firm. And this is our podcast where we go deep on all things business and technology with world-leading experts. Welcome everybody. Thank you for being here. My name is Ben Kaznoka, partner at Village Global, and on behalf of Anne and Amrit and the whole Village team, welcome to another outstanding afternoon of Village founders and friends, and of course Howie Liu from Airtable. Now Howie has raised Hundreds of millions of dollars of venture money, has built a business now doing hundreds of millions of dollars of annualized revenue. You've been at it for more than a decade. Who here is an Airtable user on a somewhat regular basis? Okay. And who here is like an enterprise customer or uses it organizationally wide? Okay, so we have some work to do to convert some of the—

Shame on the rest of you.

And who here has Airtable as a customer? Dion, okay, so we've got some, you know—

Yeah, we've got like, what's it called, the— Some give and take.

Balance of trade. Yes, exactly. Yes. Now, Howie, so almost everyone in the room has used Airtable as a user. Some of us are enterprise customers. What common misperceptions might we have about the business or the product that you want to clear up upfront?

So we've been on a very long journey, as you said. I mean, the product's been launched since 2015. I actually started the company about 3 years prior to that. So we've been around for a while. And I think what's been consistent but not necessarily understood is that we've always wanted to be more of an app platform than a collaboration tool. So when we first launched, we had to obviously pick an MVP that was simple enough to build quickly, although quickly for us was like 2 and a half years.

You spent 2 and a half years before you shipped a product?

Correct. Yeah, actually weirdly parallel timeline to some other great PLG companies like Figma and others, which we can come back to and like why I think that was a really interesting moment for creating this new grade of like desktop-rich, you know, kind of experiences that were web-based, collaborative, instantly shareable, and kind of disrupted like, you know, certain work patterns. But we spent 2 and a half years building this product. And the reason for that was this was not meant to just be like a simple project management tool or even like a spreadsheet replacement, but really like an app platform that was way easier to use. Now, the V1 looked like a spreadsheet. That was partly on purpose because we wanted to make it really intuitive. You could just start using it, didn't have to learn a new UI. But then also partly because we didn't have the time to build out the interface layer or additional views.

You didn't have the time after 2.5 years?

I mean, hey, we were working methodically.

And it makes sense. And obviously it worked out for you. You built this incredible tech infrastructure. But just on this meta point, if a founder— because you do some angel investing— if a founder came to you and said, Howie, here's my pitch, and I'm going to spend the first 3 years building it, and then I'll start shipping to the world.

Well, don't lead with that part. Talk about the big, big idea.

So what has to be true? What kinds of products lend themselves to that type of formation versus the lean startup iterative?

Yeah, I think, I mean, my belief now in hindsight, and I had a very short-lived startup before Airtable called eTax, ended up getting acquihired by Salesforce. So I only have a data point of 2, maybe like 1.2. But my belief is that either you have to have very strong conviction in the endgame and ideally even like the midgame of the company, meaning like if you're building SpaceX, you have enormous confidence that there's massive market opportunity at the end of the road, right? If you can really make a cost-effective, safe, scalable rocket, there is a market for rocket payloads that are cheap, right? And so, you know, there's a big payoff if you can do this. You also know you have durable advantages because other people are not going to go and just enter into this space really quickly. So, you know, the mid and endgame are strong, But then the, you know, kind of beginning of the game is really hard. I think Airtable looks kind of like that, where we thought really methodically about the space. I mean, the premise was really, you know, from working at Salesforce, realizing that there's a large class of business applications that really need a platform, not a solution, to be served well, as in you need to customize the data model per customer. Every customer has a different workflow, and maybe it's just a long tail use case that doesn't. warrant a really good best-of-breed vertical solution like a Workday for HRIS, right, or Salesforce for CRM. And even Salesforce itself did a really good job of serving customers, not because they built all the features for CRM, but because they built a platform that could be customized for every customer install. So that was the insight, and I felt pretty convicted based on, you know, both like working within Salesforce, doing some early customer validation, Talking to customers, understanding what are the different workflows you're doing in spreadsheets or underserved by current products to kind of validate the market. Obviously, you can never be 100%, but I felt pretty confident if we could nail the execution of a platform that could serve all these use cases with a real kind of application or relational database model, that we would be differentiated. We'd be competitively advantaged. So it wouldn't be easy. I mean, the fact that it took us so long to build it Meant that it wouldn't be something that other companies could just really quickly fast follow, right?

By the way, how many companies did you talk to when you were doing— over the 2 and a half years, how many prospective customers did you talk to?

Probably like dozens, not hundreds.

Wow.

And that was— do you think that's a lot?

It seems like a small number.

Yeah, well, I mean, maybe you're just—

What do people think?

It's like, gosh, you didn't build multiple views, you only talked to 12 customers.

Yeah, exactly.

It's kind of the lazy startup approach.

Yeah, yeah.

No, I mean, on that point, I think of it as actually— so in math, there's this concept of eigenvectors, which is where you have a really large multidimensional search space. But to reduce the dimensionality of the space, you can pick vectors or dimensions that best represent that space, right? So to make it very simple, imagine a 3D space. You have a bunch of dots throughout it. Well, find the 2D axis, the plane, that best captures, you know, the spatial coordinates of all the dots, right? If they actually kind of naturally tend towards, you know, a certain plane. And so I think the idea here is, you know, you don't have to go and collect every single dot, first of all, across every single dimension to understand, like, you know, the diversity of customer use cases you want to solve for. And I think, like, if you're picking 12, you know, frankly, even like 3, and I'll come back to, like, how we apply this concept even today to product development. But I think if you're picking like 12 customers across a wide range of industries, company sizes, or whatever, like the space that you want to, you know, kind of fit well into, it's pretty unlikely that you're going to overfit like exactly to just those 12 and that every other company that looks like them is going to be completely different or like your solution is going to be inapplicable, right? There's always kind of this, you know, healthy balance between, yeah, you want to collect more data, But I don't personally think you have to go and like, you know, get 100 customers to validate the premise if you have 12 or even like 5, you know, pretty representative ones that are not all like weird outliers that are super early adoptive and like nobody else will ever look like them.

Now in the early years of Airtable, there's a lot of product-led growth and kind of bottoms-up adoption. Raise your hand here if you have a product-led growth strategy for distribution enterprise side. So we have several. And Julie, we're gonna come to your question soon that you submitted for Howie.

Yeah.

More recently, you've been wining and dining some of the biggest CEOs in the Fortune 500, talking about wide-scale enterprise adoption. Talk through the evolution of the self-serve side of the business to enterprise, and how would you advise founders who are thinking about a PLG motion in the early days? 'Cause by the way, it is the dream. 9 VC pitches out of 10 in the enterprise today have founders saying, It's PLG, it's bottoms-up adoption, we don't have to hire salespeople. And then usually they end up hiring lots and lots of salespeople. So talk through that intellectual journey for you.

You know, I think there's pros and cons. So I wouldn't say PLG is like a categorically better growth model. I think the benefit obviously is like, yeah, you can have way better unit economics. Like you're not paying a cost of sale to acquire customers, especially if you're getting top of funnel organically. Then you don't even have to pay marketing costs, right? CAC to like acquire those customers. So you can literally have free growth. And, you know, in our case, like, had we wanted to go and run like a lifestyle kind of cash flow generative business, we probably could have operated the PLG engine like that, that product like up to well beyond $100 million in revenue scale with like maybe less than 30 people, which means like it just, you know, it could have been like a very cash generative product. In fact, we got to tens of millions of revenue. So we raised our Series C, our first unicorn round, I think right as we were crossing $15 million, but like very fast growing, like over 100%